Is 50° hot? Hah, I fooled you! I was talking about Celsius. That would be 122° Fahrenheit. That is hot.
But wait a second … what if I’m talking about what temperature to bake a cake? Well than that’s not nearly hot enough.
Okay, enough games. Let’s say we’re talking about a summer afternoon and it’s 30° C, which is 86° F. Is that hot?
Depends on who you ask
The thing about qualitative words like “hot” or “cold”, “high” or “low”, is that they’re all relative, and all subject to interpretation. That doesn’t make them bad words, or mean that we shouldn’t use them. We just need to understand what they’re telling us.
For instance when a study talks about the results of a “high protein diet”, what do they mean by “high”? In one example the cut-off for “high protein” was 20% of daily calories from protein. If you work out regularly and follow common nutritional guidance, you’ll think that number is pretty low.
A common pattern in observational studies is that they start with the data and ask themselves, “How could we group the subjects so that one group is statistically different from another group?” Nothing wrong with that, it’s how we find unexpected patterns.
But when the top of your range ends up towards the middle or bottom of what you’d find in a general population, you’d better be really careful about calling it “high”.